Introduction to Fabricators

Definition and types of “fabricator”

     A fabricator, also called a “fabber” for short, is a device that makes things automatically from computer data and raw material. Fig. 1 illustrates the concept of a fabricator. Long relegated to the domain of science fiction novels and movies, fabricators are very real today, although still quite primitive compared to the “replicator” of Star Trek fame.

     Fabricators trace their origin to the NC (numerically controlled) milling machine invented by John T. Parsons in the late 1940s. The CNC (computer-numerically controlled) mills, lathes, and related equipment that evolved are today standard fare in most modern manufacturing facilities. They are known now as subtractive fabricators because they work by removing material from a solid block to reveal the desired product inside. Subtractive fabricators are today a $10 billion industry worldwide.¹

     More modern technology has given us additive fabricators, machines that work by bonding small elements of an amorphous raw material, such as a liquid, powder, or film, to form the desired product. The first commercial additive fabricator, the SLA-1 made by 3D Systems of Valencia, California, was introduced in early 1988. After its first eight years, the burgeoning additive fabricator industry registered $176 million of product sales, or $421 million total sales including both products and services, in 1996.² Product sales are expected to reach $1 billion by around the turn of the century.

Segmentation: Industrial, studio, and personal fabricators

     The market for fabricators may be broken down by a combination of the characteristics of the machine and of the user. Large, complex, and expensive fabricators are generally used in an industrial setting. Most fabricators today, both subtractive and additive, fall into this category, called the industrial segment. Industrial fabbers range in price from approximately $60,000 to $10,000,000.

     Smaller, more easily operated, and less expensive machines are used in smaller workshops, either in a small or large business or possibly in a home hobby shop. For the more modern additive machines, they may also be found predominantly in an engineering office or design studio. We call this the studio segment. Studio fabbers range in price from about $15,000 to $50,000.

     At the lowest end of the price spectrum are machines we might call personal fabbers. Generally these are, or will be, tabletop machines, primarily designed for the hobby and educational market. They have lower output specifications than their higher-priced cousins, such as in speed, accuracy, material versatility, etc. A personal fabricator is generally priced under $10,000. There do exist today subtractive fabricators in this price range, and even as low as $1,000. While there are currently no additive personal fabricators, there is every reason to expect them to appear on the market within the next ten years.

     On the additive side, growth in unit sales of industrial fabbers is on the decline after an eight-year history. The decline in growth of the industrial segment is offset by a dramatic increase in growth of the studio segment, where unit sales grew by 190% from 1995 to 1996. The first personal fabricator may be expected to appear in the early years of the decade 2000, and is likely to cause a new spurt of rapid growth of the industry.

Figure 1.19. Some people speak of the output of fabricators as “desktop manufacturing” by analogy to the desktop publishing revolution. A fabricator uses computer data as instructions to create a 3-dimensional, solid object, which may be a model or prototype, or an actual product.

Comparison to the computer industry

     It is interesting to compare the history of the additive fabricator industry with the early years of computers. In 1960, about eight years after the introduction of the UnivAC, the first commercial electronic computer, about 2,000 computers had been sold around the world. The additive fabricator industry reached cumulative system sales of 2,243 in its ninth year, 1996. These included 1,997 industrial and 246 studio fabbers. The computers sold in 1960 were all large, complex machines which have come to be called “mainframes.” Later the computer market came to be broken down into three segments analogous to fabricators: the original mainframe (industrial) segment, the lower cost but high-powered “workstation” (studio) segment, and eventually personal computers.

Is there a market for personal fabricators?

     Many people today doubt that there will be any demand for personal fabricators. The situation was once the same for personal computers. In early 1981, while a team of hand-picked engineers was laboring on the IBM Personal Computer in a secret laboratory in Boca Raton, Florida, a front page story in the Los Angeles Times Business Section was headlined:

Home Computer Prices Fall — but Where’s the Market?

The article quoted an unnamed computer company executive as saying that the home computer “is a mislabel because there isn’t any home market.” IBM executives, comfortable in their dominance in large, expensive computers, apparently agreed; the initial sales projections for the IBM PC called for 250,000 units over a five year period. But IBM was wrong. Within three years, the company was shipping that many PC computers per month. By 1991, over 60 million IBM and IBM-compatible desktop computers had been sold. In 1992, personal computer revenues accounted for 61 percent of computer revenues of U.S. manufacturers.³

Fabber Processes of Today and Tomorrow

     A fabricator process is a method for manipulating matter and inducing it to take on a particular shape and structure, possibly in conjunction with control of certain other properties as well. We are, in 1997, at a very early stage in the development of processes to execute such magic. All of today’s fabricator processes fall into two broad families: subtractive and flat-layer additive.

Subtractive fabrication

     In subtractive fabrication, a cutting tool removes material from a solid block of raw material (workpiece) to reveal the desired shape. This may be done by:⁴

• Milling. Rotating a sharp blade in contact with the stationary workpiece.
• Turning. Rotating the workpiece in contact with a stationary or rotating, sharp blade.
• Wire EDM. Passing an eroding current of sparks across a narrow gap between a tensioned wire and the surface of the workpiece.

Flat-layer additive fabrication

     In flat-layer additive fabrication, amorphous material in the form of a liquid, powder, or adhesive film is caused to build up in flat layers in the proper sequence of shapes to make up the desired shape. This is currently done by four types of techniques:

• Selective surface curing. Curing successive layers of a liquid resin by a scanning laser or masked lamp working at the top or bottom surface of the resin.
• Selective sintering. Sintering successive layers of a meltable powder by a scanning laser.
• Pattern lamination (stack-first). Bonding and then cutting successive patterns of an adhesive film (this is actually a hybrid subtractive/additive technique).
• Deposition. Depositing successive layers of a fusible material. This in turn may be done by laying a continuous stream of material (continuous deposition), by jetting a stream of fusible droplets (drop-on-drop deposition), or by jetting a stream of binding droplets on successive layers of a powder (drop-on-powder deposition).

Improvements in flat-layer techniques

     Some alternative flat-layer techniques and improvements in existing ones are under development. Some examples include:

• Form-first pattern lamination.⁵ Forming successive patterns of an adhesive or cohesive material on a carrier sheet and then conveying the patterns on the sheet for bonding in a stack. The forming may be performed by cutting the pattern in an adhesive film (called in that case “cut-first pattern lamination,” a hybrid additive/subtractive technique) or by depositing material selectively on the carrier sheet (called “additive pattern lamination” because this version is fully additive, not hybrid).
• Alternative curing or sintering techniques, such as by electron beam instead of by laser, or by a panel of light-emitting diodes (LEDs), an array of microlasers, or a digital micromirror device (DMD).
• Alternative deposition techniques, such as depositing a stream of powder into the path of a scanning laser, inducing vapor deposition by a scanning laser, or depositing a pattern of two complementary powders in which one is and the other is not subject to fusing by a postprocess such as melting.
• Edge-controlled layer-additive fabrication. A modification of some of today’s flat-layer techniques in which a particular slope or curvature is induced in the edges of the layers. This eliminates or reduces the “stair step” effect which causes a rough surface on sloped edges of objects made by layer-additive techniques. It also allows for much faster fabrication by allowing thicker layers to be used. Edge control may be performed by angled cutting in pattern lamination, by trowelling in continuous deposition, or by layer-by-layer edge milling in those two techniques or in drop-on-drop deposition. It cannot likely be performed in techniques in which the edges are buried by powder or curable resin.

Improvements beyond flat layers

     Future fabricators will go beyond the limitations of just subtractive and flat-layer additive techniques. Some concepts which are being explored or which we can expect to be explored in time include:

• Curved-layer pattern lamination, in which the succession of adhesive patterns is laid up on a curved-surface base. This is most beneficial if the adhesive film is a long-fiber composite prepreg because it then allows the process to control the internal orientations of the fibers in order to optimize the anisotropic properties of the fabricated product.
• Freeform fabrication,⁶ in which the product is built up additively in arbitrary fashion, not constrained to a sequence of flat or curved layers. Some possibilities for achieving this are:
  • Freeform deposition. All deposition techniques that do not involve a loose powder (the drop-on-powder and two-complimentary-powder techniques) are readily adaptable to this more general method. The barrier to accomplishing it is primarily one of software control of the process since layer-by-layer fabrication is easier to calculate. The software challenge in freeform deposition is like an infinite-dimensional derangement of today’s already difficult problem of automating the selection of build orientation in any of the layer-additive processes. A drop-on-drop fabricator must be vector-style (as in the one made today by BPM) to work in freeform fashion; raster-style devices (such as Sanders’ machine and the Actua by 3D Systems) are only capable of working in layer-by-layer fashion.
  • Selective interior curing (dual-beam curing). Curing a succession of points in the interior of a vat of resin by scanning two lasers of complementary frequencies such that curing occurs only at the intersection of the beams.
• Formative fabrication. This is neither subtractive nor additive. Instead of removing or building up material, a formative process works by applying forces to opposing sides of a given mass of material in order to induce the formation of shape. This is like molding, except without the need for an expensive mold to be manufactured for each shape desired. Just as pattern lamination today is a hybrid subtractive/additive technique, we can ultimately expect to see trihybrid processes that combine the advantages of additive, subtractive, and formative techniques in a single machine.
• Augmentation by 1- and 2-D materials. If personal fabbers are ever used to make children’s toys, one of the worst problems they will have with it will be in making fur, hair, and eye lashes. Various processes will be needed to augment 3-D fabrication with manipulation of 1-D and 2-D materials and the ability to incorporate those materials into a product. Such processes might include braiding, weaving, stitching, bending, jabbing (to affix hair to a surface), and tacking.
• Accretive fabrication. Far off in the future, when additive fabrication is a mature industry and increased use is being made of the then-new formative processes, there will begin the development of a new class of techniques that follow more from biology than from any industrial process. For lack of a better term, we will call them here “accretive,” the meaning being that the product grows by adding material to itself rather than by the action of any outside influence. This implies that the control software and the process technology must both be incorporated into the elements of the product as it grows, just as is the case with the growth of living organisms by mitosis under the control of DNA. Processes that we understand today that might help in this category include secretion, agglomeration (including aided by secretion), self assembly, and morphogenesis.

The Fabricator Revolution

     Today’s subtractive and flat-layer additive processes may be seen as the first two waves of a progression of technology development that may be called the fabricator revolution, a new economic era in which material products are rendered automatically from raw materials under computer control. It is important to see today’s technologies in this light so as not to be trapped in the notion that what we have now is the whole story. What we have today is only a shadow of what is coming.

     Visionary observers of the UnivAC electronic computer in 1952 imagined future computers running much faster, communicating more naturally with people, and being smaller and more manageably operated. However, even though the transistor had already been invented, it is unlikely that anyone yet, in 1952, foresaw the specific technology of integrated circuits and microprocessors. Therefore, while it is important to speculate about what sorts of technologies might be possible, and what sorts of technologies we might be able to invent, we must realize that the best future technologies may be beyond speculation right now.

     If you found this interesting, you’ll also want to read:

• Growing Autofab into the 21st Century, based on an invited lecture at a United Nations conference in Ibadan, Nigeria, discusses the three basic areas of research in fabber development: process, materials, and control, with examples of the sorts of problems studied in each area.
• Perspectives on StereoLithography—Automated Fabrication in the 19th, 20th and 21st Centuries, Marshall Burns’ first keynote address, given to the 1992 StereoLithography Users Group Conference in San Francisco. He presented many of the exciting lessons learned in the research for his book, Automated Fabrication, which was published the following year. This included the origins of and future prospects for this dynamic industry, with suggestions on handling its challenges.

Footnotes